Global Ensemble and NAEFS Yuejian Zhu and Zoltan

Global Ensemble and NAEFS Yuejian Zhu and Zoltan Toth Environmental Modeling Center NOAA/NWS/NCEP Acknowledgements: R. Wobus, M. Wei, B. Cui, D. Hou, M. Iredell and S. J. Lord EMC B. Gorden, S. Jacobs and D. Michaud NCO E. Olenic, D. Unger and D. Collins CPC Presentation for 3 rd Ensemble User Workshop October 31 st 2006

Outlines q q q q q NAEFS History and Milestones GEFS, NAEFS and THORPEX Review GEFS Implementation (FY 06) Review First NAEFS Implementation (FY 06) Ensemble Products and Functionalities Ensemble Data Request Information GEFS Major Implementation Plan (FY 07) NAEFS Upgrade Plan (FY 07) NAEFS Expansion and Future Plan

• NAEFS History and Milestones February 2003, Long Beach, CA – NOAA / MSC high level agreement about joint ensemble research/development work (J. Hayes, L. Uccellini, D. Rogers, M. Beland, P. Dubreuil, J. Abraham) • May 2003, Montreal (MSC) – 1 st NAEFS Workshop, planning started • November 2003, MSC & NWS – 1 st draft of NAEFS Research, Development & Implementation Plan complete • May 2004, Camp Springs, MD (NCEP) – Executive Review • September 2004, MSC & NWS – Initial Operational Capability implemented at MSC & NWS • November 2004, Camp Springs – Inauguration ceremony & 2 nd NAEFS Workshop • Leaders of NMS of Canada, Mexico, USA signed memorandum • 50 scientists from 5 countries & 8 agencies • May 2006, MSC & NWS – 1 st Operational Implementation • Bias correction • Climate anomaly forecasts • June 2006, Montreal (MSC) – 3 rd NAEFS Workshop • March 2007, 2008, MSC, NWS – Follow-up implementations-Improved and expanded product suite

GEFS, NAEFS and THORPEX • NCEP Global Ensemble Forecast System (GEFS) is part of NAEFS • NAEFS is combining NCEP and CMC global ensemble • THORPEX is the research project: – Provides framework for transitioning research into operations – Prototype for ensemble component of THORPEX legacy forecast system: Global Interactive Forecast System (GIFS) THORPEX Interactive Grand Global Ensemble (TIGGE) Transfers New methods Articulates operational needs North American Ensemble Forecast System (NAEFS)

Review GEFS Implementation (FY 06) 1. Increase the number of perturbed ensemble members • • 14 (in place of current 10) perturbed runs for each cycle (20 by early 2007) • NAEFS requirement This change is intended to improve ensemble based prob. forecasts • Results: improved probabilistic skill, slightly improved ensemble mean skill (seasonally dependent) 2. Add control runs for 06, 12 and 18 Z cycles • • This change is intended to enable for relocation of perturbed tropical storm Facilitates comparison of high & lower resolution ensemble controls • If lores control and ensemble mean differ – indication of nonlinearities • If high & lores controls differ – indication for possible effect of resolution 3. Introduce Ensemble Transform (ET) into GEFS breeding method • • • ET breeding method creates globally orthogonal initial perturbations Uses simplex method to create individual (not paired) perturbations This change is intended to improve probabilistic forecast skill • Results: Improved probabilistic forecast skill; Slightly reduced ensemble mean hurricane track errors for 12 -96 hours 4. Changes of File Names and Structures

GEFS configurations Current Plan Model GFS Initial uncertainty BV ETBV Model uncertainty None Tropical storm Relocation same Daily frequency 00, 06, 12 and 18 UTC same Hi-re control (GFS) T 382 L 64 (d 0 -d 7. 5) T 190 L 64 (d 7. 5 -d 16) same Low-re control (ensemble control) T 126 L 28 (d 0 -d 16) 00 UTC only T 126 L 28 (d 0 -d 16) 00, 06, 12 and 18 UTC Perturbed members 10 for each cycle 14 (20) for each cycle Forecast length 16 days (384 hours) same Implementation August 17 th 2005 May 30 th 2006

Ensemble Transform Bred Vector (Plan) Bred Vector (Current) Rescaling ANL P 1 Rescaling P 1 forecast P 2 forecast ANL N 1 P 3 forecast t=t 0 t=t 1 t=t 2 P 4 forecast t=t 0 t=t 1 t=t 2 P#, N# are the pairs of positive and negative P 1, P 2, P 3, P 4 are orthogonal vectors P 1 and P 2 are independent vectors No pairs any more Simple scaling down (no direction change) To centralize all perturbed vectors (sum of all vectors are equal to zero) P 2 Scaling down by applying mask, The direction of vectors will be tuned by ET. ANL N 2

Changes of File Names and Structures • Pressure GRIB Files Split into Two – Pgrba – 51 Variables For NAEFS Exchange – Pgrbb – Remaining 278 Variables • Perturbation runs from pairs to single size – P 1, n 1, p 2, n 2 … convert to p 0, p 02, p 03, p 04 … • Enspost. * and Ensemble. * Files Eliminated – Data Was Duplicate to Pressure GRIB Data Packed in Different Format • Ensemble Extensions Corrected in Pressure GRIB Files • 6 -Hourly Precipitation/Max/Min Accumulations Available in Pressure GRIB Files • GEMPAK Files Created for NAEFS Members for Raw and Bias Corrected pgrba Files • GEMPAK Metafiles For HPC Medium Range Desk Created in Production

Review First NAEFS Implementation (FY 06) 1. Bias corrected members of joint MSC-NCEP ensemble • Decaying accumulated bias (~past 50 days) for each var. for each grid point • • • For selected 35 of 50 NAEFS variables 32(00 Z), 15(06 Z), 32(12 Z) and 15(18 Z) joint ensemble members Bias correction against each center’s own operational analysis 2. Weights for each member for creating joint ensemble (equal weights now – unequal weights to be added later) • • • Weights don’t depend on the variables Weights depend on geographical location (low precision packing) Weights depend on the lead time 3. Climate anomaly percentiles for each member • Based on NCEP/NCAR 40 -year reanalysis • • • Used first 4 Fourier modes for daily mean, Estimated climate pdf distribution (standard deviation) from daily mean For selected 19 of 50 NAEFS variables 32(00 Z), 15(06 Z), 32(12 Z) and 15(18 Z) joint ensemble members Adjustment made to account for difference between oper. & re-analysis Provides basis for downscaling if local climatology available – Non-dimensional unit

Schematic diagram forecast anomalies Climatology Medium forecast (50%) Bias-corrected Modified climatology Ensemble forecast Temperature Lower extreme Upper extreme (10%) (90%) Clmatology is generated from NCEP/NCAR reanalysis (40 years from 1958 to 1997)

ENSEMBLE 10 -, 50 - (MEDIAN) & 90 -PERCENTILE FORECAST VALUES (BLACK CONTOURS) AND CORRESPONDING CLIMATE PERCENTILES (SHADES OF COLOR) Example of probabilistic forecast in terms of climatology

Ensemble Product Request List NCEP SERVICE CENTERS, OTHER PROJECTS

Ensemble Functionalities List of centrally/locally/interactively generated products required by NCEP Service Centers for each functionality are provided in attached tables (eg. , MSLP, Z, T, U, V, RH, etc, at 925, 850, 700, 500, 400, 300, 250, 100, etc h. Pa) FUNCTIONALITY CENTRALLY GENERATED 1 Mean of selected members Done 2 Spread of selected members Done 3 Median of selected values Done Sept. 2005 4 Lowest value in selected members Done Sept. 2005 5 Highest value in selected members Done Sept. 2005 6 Range between lowest and highest values Done Sept. 2005 7 Univariate exceedance probabilities for a selectable threshold value Done, Dec 05 8 Multivariate (up to 5) exceedance probabilities for a selectable threshold value Done, Dec 05 9 Forecast value associated with selected univariate percentile value Done Sept. 2005 10 Tracking center of maxima or minima in a gridded field (eg – low pressure centers) Done Sept. 2005 11 Objective grouping of members Planning starts FY 06, Deliver FY 07 -08 12 Plot Frequency / Fitted probability density function at selected location/time (lower priority) Detailed Planning FY 06, Deliver FY 07 13 Plot Frequency / Fitted probability density as a function of forecast lead time, at selected location (lower priority) Detailed Planning FY 06, Deliver FY 07 14 Spaghetti (ability to interactively change contour/domain etc) Basic function done; Interactive version to be scheduled (TBS) Additional basic GUI functionalities: - Ability to manually select/identify members (TBS) - Ability to weight selected members Done, Sept. 05 LOCALLY GENERATED INTERACTIVE ACCESS Potentially useful functionalities that need further development: - Mean/Spread/Median/Ranges for amplitude of specific features (TBS) - Mean/Spread/Median/Ranges for phase of specific features (TBS)

Ensemble Data Request Information • AT NCDC : Over 10 days, 09/26 -10/05 0026 0003 1527 1860 meteo. noa. gr dip 0. t-ipconnect. de weathersa. co. za fsu. edu • In a ten day period 09/26 -10/05 0001 retail. telecomitalia. it 0001 196. 12. 132. 227 0001 rr. com 0002 proxy 2. enpc. fr 0003 202. 131. 2. 222 0065 fi. upm. es 0071 bruneiweather. com. bn 0150 abo. wanadoo. fr 0152 meteo. noa. gr 0297 zedxinc. com 0968 buran. meteotest. ch 1050 nps. edu 1141 nuvox. net 1298 sd. cesga. es 1927 live-servers. net 9489 psu. edu 137069 cox. net

Ensemble Data Request Information (Cont. ) • A list of the users who pulled GEFS data from the NCEP anonymous ftp server in September, and have left a valid email address at login. – 26 e-mail lists • A list of the users who pulled GEFS data from the WOC server in the past month, and have left a valid email address at login. – 26 e-mail lists.

Ensemble Web-Page Access Information Usage Statistics for GMB ENS www. emc. ncep. noaa. gov Summary Period: Last 12 Months Generated 30 -Oct-2006 14: 30 EST

GEFS Major Implementation Plan (FY 07) • Upgrade vertical resolution from 28 to 64 levels for 20 perturbed forecasts – 4 cycles per day – T 126 L 64 – Up to 384 hours (16 days) • Real-time generation of hind-cast at T 126/L 64 resolution. – – 4 cycles per day 27 hind-casts for each cycle since 1979 Using reanalysis II initial conditions (T 62 L 28 resolution) Add random noise to high frequency (T 63 -T 170) by using • Cycling (6 -hr T 170 model forecast) • Other method? • (Alternate) – upgrade both horizontal and vertical resolution to T 170/L 64 • Introduce ESMF scheme that allows concurrent generation of all ensemble members. • Add stochastic perturbation scheme to account for model errors (tentative plan)

NAEFS upgrade plan (FY 07) • Add approximately 15 new variables to current 51 pgrba for NAEFS data exchange. – Such as vertical shear, helicity, u, v, t, RH for 100, 50 h. Pa, LH, SWR, LWR at surface, and etc. . • Add GFS high resolution control bias correction by using current method for ensemble. – There is a problem when we estimate bias after GFS change resolution after 180 hours • Set up GFS low resolution (ensemble) control run on NCO’s real time parallel prior to GFS upgrade in the future. – As bias estimation of GFS major/minor implementation – Need to compare the bias of ensemble mean and control • Improve bias correction algorithm. – Pending on hind-cast information – Two weights: one from real-time (analysis and forecast) bias estimation (mainly for week-1), another one from hind-cast (mainly for week-2)

NAEFS Expansion and Future Plan • Plans to be coordinated with THORPEX – Links with Phase-2 TIGGE archive and beyond (GIFS) • Expansion – FNMOC • Experimental data exchange by Dec 2006 • Preliminary evaluation by Dec 07 • Operational implementation by Dec 08 (subject to improved performance) – UK Metoffice • Decision on going operational & possibly joining NAEFS - by 2008 – KMA, CMA, JMA • Expressed interest, no detailed plans yet • Data exchange with MSC – Replace current ftp with more reliable telecom by Dec 08 • Statistical post-processing – Continual enhancements to current methods (2 nd moment correction, addtnl vars) – Testing (Dec 08) & possible implementation (09) of advanced methods • Products – Week-2 – experimental by Nov 06 – Web graphics • MSC – Nov 06 • NCEP – Mar 07 19

Background !!!!!

Bias Correction Method & Application § Bias Assessment: adaptive (Kalman Filter type) algorithm decaying averaging mean error = (1 -w) * prior t. m. e + w * (f – a) For separated cycles, each lead time and individual grid point, t. m. e = time mean error 6. 6% • Test different decaying weights. 0. 25%, 0. 5%, 1%, 2%, 5% and 10%, respectively 3. 3% 1. 6% • Decide to use 2% (~ 50 days) decaying accumulation bias estimation Toth, Z. , and Y. Zhu, 2001 § Bias Correction: application to NCEP operational ensemble 15 members

List of Variables for Bias Correction, Weights and Forecast Anomalies for CMC & NCEP Ensemble

Recently Statistics

Based on raw forecasts, no climate and current analysis correction